Rubbish Agro; I have linked to Spencer's revised graph above where he removes the base period; here it is again:

http://www.drroyspencer.com/wp-content/uploads/CMIP5-73-models-vs-obs-20N-20S-MT-5-yr-means1.png

UAH does not use a base period; it establishes anomalies based on the average of ALL its data. This removes the real bias inherent in a base period which you gloss over.

Your comments about the Troposphere are therefore spurious. The absence of a THS remains a fundamental and fatal flaw in AGW theory.

As for ground based temperature; they have not risen for near climatically significant periods. Werner Brozek has proven this by using temperature data and applying the criteria used by NOAA and Jones at UEA CRU. Even using the standards and methods of AGW science itself disproves it; see point 1 here:

http://theclimatescepticsparty.blogspot.com.au/2013/06/man-made-global-warming-wrong-ten.html

But that simple fact, that AGW theory has been disproved, won't stop you will it Agro?

Which humanities subject is your PhD in again? English poetry? Communications? Journalism?

Anthony, perhaps you should read that graph again. You know the bit where it states: “Tropical Mid-Troposphere Tropical 20S-20N” again. Then you can tell me how that is a revised graph of a comparison of global surface temperatures with models.

I think this fully illustrates the depth of your understanding. You somehow think that tropical mid-troposphere temperatures are interchangeable with global surface temperatures.

You have for about the fifth time misunderstood my complaint about Spencer’s graph. I did not make any statement about the base period used to generate UAH anomalies. My complaint was that Spencer used the UAH period of 1979-1983, when UAH over-estimates global temperatures, as the base for their comparison between models and data. I am uncertain whether your continual confusion over this is simply ignorance or if it is deliberate.

Oh Anthony, that silly blog post of yours. Gosh it gave me the giggles. Starting the trend lines at 2004.75 or 2001.33 or 2000.9. Why? How long did you spend having to cherry pick start points to get all the trend lines to look flat? You really don't understand this science stuff do you.

Agro; my silly little blog post is based, as is Werner's findings of statistically significant pauses in temperature on land, on the criteria offered by NOAA, CRU and Santer. Did you bother to even read Werner's methodology which is the methodology of these official sources to establish trends. There is no cherry-pick; the lengths of a pause in the data of each temperature indice is based on those 'official' criteria applied to the data to establish the length of the pause!

This is hopeless, you are so immersed in the paradigm your intellect, considerable no doubt, might as well be putty.

If a theory, sorry alleged theory, like AGW is contradicted by its own methods and criteria and that contradiction is still not accepted by its acolytes then what?

If you had also bothered to understand my post I did distinguish between Troposphere temperature and surface temperature which is why I used Spencer's graph of the Troposphere and Werner's analysis of the surface.

If you wasted less time being a smartarse you might be able to meaningfully contribute.

Oh Anthony, some mothers do ‘ave ‘em, don’t they?

Just because a retired teacher embarrasses themselves on-line, doesn’t mean you have to follow suit. The problem with your graph is that you have worked back from today to find the last time in the data sets when the slope of the line is close to 0 and because one of them is more than 15 years concluded it has not warmed in 15 years. This is simply cherry picking. It is a conclusion looking for data.

The proper way to do this is to deal with the hypothesis. If that hypothesis is that there has been no warming for the last 15 years, then you test the last 15 years, not 8 years, not 12 years, not 13 years, not 16 years, but 15 years.

Like this: http://www.woodfortrees.org/plot/hadcrut4gl/from:1998.75/to:2013.75/trend/plot/uah/from:1998.75/to:2013.75/trend/plot/gistemp/from:1998.75/to:2013.75/trend/plot/rss/from:1998.75/to:2013.75/trend

Three out of 4 data sets show a clear warming over the period. Even better, I have gone to the original data for these three (I don’t have the RSS data set, but I am guessing that would not be significantly different to 0 over the period) and done the regression and all three have a significant increase in temperature (p = 0.0027 for Giss; p = 0.048 for HadCRUT4; p = 0.044 for UAH).

Regardless of what you have done with Brozek’s graph, it doesn’t in any way address the criticism of Roy Spencer’s graph.

Agro, Spencer's graph, used in this article, applies to both land and Troposphere temperatures; I have supplied Spencer's graph corrected for your complaint for the Troposphere where there is no short base period to calculate anomalies but where the whole data is used to produce anomalies.

Let me ask you directly: do you have a complaint about that method?

As for Werner Brozek's land and satellite temperature analysis. Apart from your snide aspersion on teachers your comment about his method misses the point.

What is wrong in back-searching the data on a monthly basis to find a negative slope and then seeing if that persists for the remainder of the consequent data?

You also speak of a "a significant increase in temperature (p = 0.0027 for Giss; p = 0.048 for HadCRUT4; p = 0.044 for UAH)."

This is why I don't take you seriously. As well as looking at periods of negative slope, that is declining temperature, Brozek also looks at periods of increase in temperature with statistically insignificant periods of increase. What does that mean? It means the increase is not statistically different from zero when error bars, or the margin of error, are used. Error bars reflect the possibility of error in the data and method. Brozek has used the data and error bars from Skeptical Science.

How much fairer could he be?

So when we look at your conclusion of a significant increase in temperature, what you mean is that the temperature increase is statistically real but when the error bars are considered it is not statistically significant. That is why your P values are LESS than 0.05, the usual threshold value for determining statistical value.

What would be appropriate would be to return to Brozek's starting point of negative slope and see whether the result is maintained given Brozek did his analysis in April of this year, 6 months ago.

Oh dear Anthony, you are such a chump.

Spencer’s graph given by Don Aitkin was for the surface and the lower troposphere (you know that bit of the atmosphere closest to the Earth). It says so on the graph. It also says on the graph that it is for global temperatures. The graph you linked to as a ‘correction’ was for the mid-troposphere (a bit of the atmosphere not that close to the Earth) and for the tropics. It says so on the graph.

It also can’t be a correction for the graph posted by Aitkin, because it was published on June 6 http://www.drroyspencer.com/2013/06/still-epic-fail-73-climate-models-vs-measurements-running-5-year-means/ when the graph it was apparently correcting was published on October 14 http://www.drroyspencer.com/2013/10/maybe-that-ipcc-95-certainty-was-correct-after-all/ some 4 months later – unless Spencer among his other gifts is able to time travel.

I have suggested before that you should keep away from commenting on statistics due to your thorough ignorance of the area. You will only end up looking like a fool, as this example shows:

“So when we look at your conclusion of a significant increase in temperature, what you mean is that the temperature increase is statistically real but when the error bars are considered it is not statistically significant. That is why your P values are LESS than 0.05, the usual threshold value for determining statistical value.”

I really don’t know what to say.

Perhaps you should read up a little http://www.wikihow.com/Assess-Statistical-Significanc